ABSTRACT Alcohol and marijuana (AM) are the most commonly used substances among adolescents in the U.S. The consequences of AM use are significantly higher relative to use of either substance alone. Teens? propensity to engage in multiple risky behaviors while driving combined with their relative inexperience with the timing and duration of marijuana?s effects puts them at heightened risk for experiencing harms related to driving under the influence. Marijuana use initiation and binge drinking peak at age 16, the legal age teens may apply for a provisional driver?s license in some states. Thus, targeting these novice teen drivers prior to licensure is an ideal time for prevention efforts focused on reducing alcohol and/or marijuana initiation, use, and impaired driving. This study builds on effective interventions that have demonstrated reductions in alcohol and/or marijuana use and reduced consequences one year later, and proposes to adapt one of those interventions, CHAT, to the web (web-CHAT). We will evaluate the efficacy of web-CHAT among 15.5-16-year-old adolescents (n=150) recruited when they are attending behind-the-wheel training at the Automobile Club of Southern California (AAA). The study has the potential to promote public welfare by improving adolescent health outcomes and reducing risky driving behaviors that can have substantial monetary and social costs, as well as by providing unique insight into what mediates reduced risky driving attitudes behaviors among those in the intervention. The study is innovative because it is for both youth who are at risk for substance use as well as those who are not, and it is delivered during a teachable moment when adolescents receive driver?s education. We also collaborate with a leading provider of youth driver education, AAA, which provides online curriculum to AAA and non-AAA programs across the country. Finally, our study can provide unique insights about the efficacy of web-CHAT to reduce marijuana initiation, use, and risky driving attitudes in the context of a changing marijuana policy climate. We propose a 3-year study to test the feasibility of our procedures in a driver education setting and pilot the efficacy of web-CHAT. We will test whether web-CHAT reduces alcohol and/or marijuana initiation or use, and reduced risky driving attitudes and behaviors (intent to drive after drinking/using, passenger with someone who drank/used) compared to teens in UC, at three and six-month follow-up. We propose a crucial, time-sensitive intervention during driver education. This proposal allows us to establish the feasibility of our recruitment and study workflows, while also establishing acceptance and efficacy of web-CHAT among a diverse group of adolescents.